Concept & Prototype Testing Quick Guide — The UX Research Library
The UX Research Library

Concept & Prototype Testing Quick Guide

A table read for your idea, a dress rehearsal before it ships. Pick a section — worth bookmarking for next time.

Tap a section to jump straight to it

Sketch or Prototype?

The table read comes before the dress rehearsal — don't skip ahead.

Test a sketch (the table read) when
  • You have more than one direction and need to know which is worth building further.
  • The idea is still cheap to throw away — you want reactions before you've fallen in love with it.
  • You need to know if the concept makes sense at all, before worrying about how it works.
Test a prototype (the dress rehearsal) when
  • The concept is settled and now you need to know if people can actually use itPrototypes answer "does this work," not "is this the right idea" — testing one too early wastes fidelity you didn't need yet.
The costly mistake
  • Building a polished prototype before testing the underlying conceptA beautiful prototype of the wrong idea is still the wrong idea — just a more expensive version of it.

Who to Test With

Casting the right audience matters more than casting a lot of them.

Good enough for early sketches
  • 5 people who roughly match your audience is plenty at this stage — you're looking for a pattern, not a statistic.
  • Colleagues outside the project can work for a first gut-check, as long as you're clear that's what it is.
Get pickier for prototypes
  • By prototype stage, testers should genuinely match your real audienceYou're now testing usability for real users, not just reaction — a mismatched tester's confusion may not mean anything.
Skip this group entirely
  • Anyone who already knows what the concept is "supposed" to doPrior knowledge fills in gaps a real first-time user won't have — their success doesn't predict a stranger's.

Running the Session

A grab-and-go bank of task framings — come back to this before every session.

For a sketch (react, don't perform)

"Walk me through what you think this is, before I explain anything."

"What would you expect to happen if you tapped/clicked here?"

"What's confusing, even a little, about what you're looking at?"

"If a friend asked what this does, what would you tell them?"

For a prototype (task, don't guide)

"Your goal is to [specific outcome] — show me how you'd try."

"Keep talking out loud, even when it feels obvious what to do."

"What would you do next, and why?"

"Pretend I'm not here — what would you try if you got stuck?"

Reading the Signal

Early reactions are quiet. Here's how to actually hear them.

Real signal
  • Hesitation before acting — even a pause of a second or two is worth a follow-up question.
  • Someone re-reading the same part twice without you prompting them to.
  • "Oh, I see" said in a tone that doesn't match the words.
Weak, easy-to-overrate signal
  • "I like it" with nothing specific attachedPolite enthusiasm is the most common false positive in early testing — always ask "what specifically" before counting it.
Don't trust this at all
  • Whether they say they'd use it in the futureStated future intent is famously unreliable — watch what they do with the thing in front of them right now, not what they predict about tomorrow.
The one thing to remember

What they hesitate on matters more than what they compliment.

Part of The UX Research Library — leiamanin.com